January 20, 2012
Yahoo! Calendar and iCal with Caldav fails – a fix!
Over the last few weeks I’ve had odd errors with Yahoo! calendars and Caldav with Apple’s iCal on Snow Leopard. Sometimes it refuses to log in, sometimes it logs in but there are no calendars listed, and sometimes there are no appointment updates.
Three things seem to help:
- Deleting the calendar cache
- Deleting the iCal preferences
- Adding an appointment to Yahoo! Calendar with the web interface
The last one is interesting – I mainly use the Yahoo! calendar to access my family’s diary, which is only updated from mobile devices like an iPhone using Caldav. It seems that Yahoo! goes to sleep if the web interface isn’t used occasionally. Logging on from a browser and adding an appointment kicks it into life. You can delete the appointment afterwards…
November 10, 2011
Using Google Talk without a gmail account
Normally, Google makes using its tools pretty easy. Until last night when I needed to get on a Google Talk session with some support guys in the USA and Spain and needed a Talk login pretty fast.
Kindle – problem connecting to wifi network
I guess I haven’t posted for a while 
We got our son a Kindle for his birthday, but it hadn’t occurred to me that it has a browser (or games, but that’s another story). So bedtime reading became bedtime browsing, which was not the idea!
June 10, 2011
IPv6 – 5 top tips for your business
Top 5 things you must do about IPv6
Let’s get to the point – just because the world is running out of new IP addresses to give out doesn’t mean the internet will stop. But what DOES it mean, and how does it affect the IT services you provide to your company? What exactly should you DO?
1. Check your internet-facing applications
If you run a website, an e-commerce application, a VPN appliance – in fact, any internet facing system – you need to check if it has any reliance on the IP address of the client systems. An example would be an e-commerce system that tries to work out the customer’s location from their IP address, then switches to the appropriate currency and shipping rates. This won’t work with IPv6, so you’ll need to find another way to solve the problem such as asking customers to select their location. For other incompatibilities you, or your software vendor, will need to update your application.
2. Check how well-connected your hosting provider is in the IPv6 world
In the old IPv4 world, the biggest and best-connected service providers are known as Tier 1 providers. They’re so well-connected that they sit like the troll under the bridge, charging everyone else to cross their networks. In the new IPv6 world though, these providers don’t have the same connectedness as they used to. In fact, they have largely been replaced by mid-level players like Claranet – big enough to invest in an IPv6 enabled network and to peer with others, small enough not to have the legacy of the old guard.
But how do you tell whether your chosen provider is well connected? The easiest way is to ask them, and ask for evidence. If you’re feeling brave, try to figure out the rather technical output of the website https://www.sixxs.net/tools/grh/dfp/all/?missing=8426 using the tool “Missing DFP’s per GRH participant”, which lets you check how many IPv6 routes are missing from your provider’s routing table. Claranet’s customers can be reached from over 98% of the IPv6 internet!
3. Ask your provider about their stock of IPv4 addresses
Managed service provides who have been careful about managing their IPv4 address allocations should have enough to allow you to grow your business, but you should talk with them about your plans as early as you can, and maybe even reserve addresses. You only need to do this for systems like web sites, firewalls and load balancers that are connected directly to the net, and with support from your provider you can make sure you’re using address conservation techniques like NAT to best effect.
4. Understand what happens to your office desktop PCs
You don’t have an IPv6 network, right? Wrong! Windows Vista has IPv6 enabled by default, as does Windows 7. Sadly, Microsoft’s implementation is a bit broken (see here http://www.zdnet.com/blog/networking/fixing-windows-7-ipv6-headaches/257) but the point is that you probably have it on your network today. Mac users can’t be smug, since they have it too. This isn’t a problem in itself, but you have to understand a bit about how v6 works to see where it can trip you up. For more background you could read this http://www.circleid.com/posts/20101012_ipv6_and_the_fear_of_brokenness/.
When your desktop tries to resolve a name to an IP address, it uses a system called Domain Name Services which tells your PC what IP address you need to use to reach the web server you want to get to. It’s like directory enquiries getting you a phone number for a store. If the web site is v6 enabled, your PC will get the IPv6 address for that site first. Then, since the PC knows about v6, it will try to use it to contact the web site. But, you don’t have an IPv6 network connection to the internet! So the traffic gets dropped, and the PC eventually falls back to using the old IPv4, but your user has seen a delay and complains that “the internet connection is slow”.
There is another scenario, thanks to Microsoft creating a translation mechanism called Teredo. Windows clients with IPv6 addresses (Vista and 7 auto-assign these by default, remember) can contact machines on the internet called Teredo relays, which help to encapsulate v6 traffic in v4 packets, then pass that traffic on to the destination server. There are some startling implications in this, not least of which is the way this can punch holes through your firewalls and content filters.
My recommendation for business users – turn off IPv6 on Microsoft systems until you’re ready to deploy it in a planned way, or get your firewall administrator to block Teredo and other tunnelled traffic!
5. Get ready for the transition!
On 8th June 2011 a group of major content providers turned on IPv6 for their servers, to give the v6 world a global scale test-flight. In other words, they switched it on to see what happened! You can test your Internet connection today here <http://test-ipv6.com/>. In the coming months, participating organizations will be working together to publish help guides with more specific instructions for diagnosing and addressing potential issues.
March 26, 2011
Education and experience: is there an ideal mix for becoming a CIO?
And again, another article quoting me!
http://www.computing.co.uk/ctg/analysis/2027217/education-experience-ideal-mix-cio
CTO v CIO: Which path should you take?
How lovely – I’ve been quoted in Computing Magazine!
http://www.computing.co.uk/ctg/analysis/2036075/cto-cio-path
December 10, 2010
Top 5 tips for recruitment companies
So here I am, now CIO for an ISP/managed hosting company, and I occasionally recruit staff. It’s never an easy process – apart from finding the right skills, I have to find the right cultural fit, the right attitude and so on. Some recruiters make this easy and do a great job. Those are the people I use. I’ve used them across 2,3 or even 4 companies, recruiting all sorts of technical, BA and PM roles.
Inevitably, I get cold-called from my LinkedIn profile by other recruiters, or from times when I’ve contacted them – maybe to apply for a role myself. It’s these companies who are the target for the advice I give freely here.
Do comment and let me know your experiences too!
So, to my pearls of wisdom:
- If you got my name from your CV database, and your records show you didn’t call me back, or email me back, or respond to my enquiry, DO NOT BOTHER TO CALL ME. You are on my “do not deal with” list. Contrary to what your ego is now telling you, this is not personal pique. If that’s how you treated me as a candidate, that’s how you’ll treat the candidates for roles I’m recruiting. I don’t want to be tarred with your brush.
- If you got my name from LinkedIn, and my profile doesn’t say that I’m recruiting and inviting CVs from recruiters, I will just delete your email. If you find my number and call, I will politely listen to you, then ignore anything you’ve said. This is simply because if you can’t read my LinkedIn profile, you can’t read a job spec or a CV. While I’m at it, don’t link to be on LinkedIn until we’ve done business. It’s even more desperate than people with 10,000 Facebook “friends”.
- When you ask me what my company does, you’ve just ruled yourself out. There’s a link to my company web site on LinkedIn. Use it. Trust me, you won’t sell anything to anybody unless you understand what they do and what they’re likely to be looking for.
- So you specialise in technical recruiting? Please learn why Solaris is not the same as Ubuntu, or why MySQL is not Postgres (or worse, SQL Server). Oh, and a CCNA doesn’t mean a candidate “knows all about networks”, sorry. You earn your fees by understanding this and filtering CVs, not by being a mail relay.
- If you send me a six line email with over 900K of clever graphical signature, it’s getting deleted. I have one in my inbox now. Click. Now I don’t. That was strangely satisfying.
So that was five tips. Here’s a bonus tip. Bet you won’t read it….
I worked for a great manager in my last role. His mantra was that the number one job of any manager is to recruit the right team, not just a bunch of skilled individuals. As a recruitment company, if you haven’t met me or my hiring managers, you have NO hope of sending the right candidates to me. THAT’s why I have built long-lasting relationships with a limited number of recruiters, and unless you can be bothered to figure our my company, my team and me you won’t get on my list of trusted partners.
Because the guys I use are just that.
October 1, 2010
Zyxel P880HW adsl router tips – nat loopback, WINS server in DHCP
Just a quick note today – I recently intalled a Zyxel P880HW ADSL router to replace my Thomson BeBox, and found a couple of things that can’t be done through the web interface.
NAT Loopback
Telnet to the router, using the same username and password as in the web interface. Enter the command:
ip nat loopback on
then, to make it stay across reboots:
type "sys edit autoexec.net" press "i", then type "ip nat loopback on" press "x" to save the configuration.
It works a lot better than the Thomson as well, which is a bonus.
WINS server setting for DHCP clients
Again, telnet to the route. Enter:
ip dhcp <interface> server winsserver <wins-ip1> <wins-ip2>
The interface is probably enif0
May 26, 2010
iPhone 3Gs battery life suddenly gets worse!
After a few years I’ve gone back to Apple products with an iPhone 3Gs 32Gb model. However, after a few weeks of using it happily with plenty of battery life for browsing, wifi, 3G and phone calls the battery life suddenly dropped to a couple of hours!
I googled for answers, most suggestions focussing on turning off features such as wifi, 3G, bluetooth and push notifications. I tried all of these with no success. I also tried deleting all the network settings and reconfiguring. No joy.
The answer?
I’m using a Microsoft Exchange account with Zarafa and z-push 1.3. For some reason the iPhone decided to constantly try to retrieve mail. Simply deleting and re-adding the mail account cured it! Now I have a full day and more of battery life.
Hope this helps someone else!
March 29, 2010
The drawbacks of Agile vs Waterfall
So all of us in tech-land have been very keen on Agile development recently. I even presented at a conference on the subject. So what’s in it for those of us responsible for delivering software solutions – and when should we decide when NOT to use it?
What’s in it for me?
From the delivery side of the deal, Agile helps a great deal. For a start, there’s much less pesky requirements gathering, arguments over price (I’ll come back to fixed-price contracts for Agile soon!), and constant moving of delivery timescales. The relationship with the customer, internal or external, is very close and the solution development and testing becomes a true team effort. I’m quite a fan of Test Driven Development, since customers are usually more interested in answering the question “what will it do” than the somewhat more abstract “what are your requirements” – the latter seems to result in long wish lists rather than maintaining a true focus on business benefits. Project scope and it’s attendant success factors are also important for management confidence.
And the drawbacks?
Without the discipline of defining requirements, whether in the form of a specification, user stories or tests, the subject matter experts from the client can fall into the habit of attempting to provide every possible feature – and it’s very hard for us suppliers to rein this in. This can result in escalations to the client management which, while being good stakeholder management practise, can take a long time while the detail is explained, and upset the SMEs in the process.
So…
In my (very) humble opinion, outsourced development could;
Use Agile when:
- The SME in the client is authoritative, and ideally has budget and delivery responsibility
- The “must have” feature list for the first release is small
- A fixed price/fixed delivery is not required
- The customer’s senior management understand what Agile is, and what is is not
And don’t use Agile when:
- A fixed price and fixed delivery are required
- The customer knows exactly what they want
- The customer’s management needs to signoff on spec, price, timescale and change control
- The SMEs are the functional experts, but not aware of the overall business processes and practises
- The company has a variety of regional processes, managers, budgets etc.
Hope this helps someone!
